DYNAMIC RESPONSE OF BRAIN SUBJECTED ... - Semantic Scholar
December 31, 2011 9:22 WSPC-255-IJAM S1758-8251 00124
International Journal of Applied Mechanics Vol. 3, No. 4 (2011) 803–823 c Imperial College Press DOI: 10.1142/S175882511100124X
DYNAMIC RESPONSE OF BRAIN SUBJECTED TO BLAST LOADINGS: INFLUENCE OF FREQUENCY RANGES
MEHDI S. CHAFI∗ , SHAILESH GANPULE∗ , LINXIA GU∗,†,‡ and NAMAS CHANDRA∗ ∗Department
of Mechanical and Materials Engineering University of Nebraska-Lincoln Lincoln NE 68588-0656, USA
†Nebraska
Center for Materials and Nanoscience Lincoln NE 68588-0656, USA ‡[email protected]
Received 31 December 2010 Accepted 21 August 2011 Blast wave induced a frequency spectrum and large deformation of the brain tissue. In this study, new material parameters for the brain material are determined from the experimental data pertaining to these large strain amplitudes and wide frequencies ranging (from 0.01 Hz to 10 MHz) using genetic algorithms. Both hyperelastic and viscoelastic behavior of the brain are implemented into 2D finite element models and the dynamic responses of brain are evaluated. The head, composed of triple layers of the skull, including two cortical layers and a middle dipole sponge-like layer, the dura, cerebrospinal fluid (CSF), the pia mater and the brain, is utilized to assess the effects of material model. The results elucidated that frequency ranges of the material play an important role in the dynamic response of the brain under blast loading conditions. An appropriate material model of the brain is essential to predict the blast-induced brain injury. Keywords: Brain; hyper-viscoelastic material model; high frequency; finite strain; blast wave.
1. Introduction Complete understanding of mild traumatic brain injuries (TBI) induced by blast waves is challenging, due to the fact that currently no medical diagnostic tools could indicate the onset of the ailment [Hoge et al., 2008]. Finite element (FE) modeling has been widely used to predict the blast-induced brain responses and better understand the mechanism of TBIs [Moore et al., 2009; Moss et al., 2009; Taylor et al., 2009; Chafi et al., 2010; Ganpule et al., 2010; Grujicic et al., 2010]. The numerical predictions depends on the appropriate material characterization under blast loading conditions. ‡ Corresponding
author. 803
December 31, 2011 9:22 WSPC-255-IJAM S1758-8251 00124
804
M. S. Chafi et al.
Brain material is strain- and frequency-dependent [Bilston et al., 2001]. Blast scenarios require knowledge of brain tissue behavior over a large strain/high frequency range [Pervin and Chen, 2009]. Current experimental research has focused on the large strain/low frequency or small strain/high frequency behavior of the tissue, as summarized in Table 1. Some material models are based on hyperviscoelastic assumptions, i.e., a linear viscoelastic model in conjunction with a nonlinear hyperelastic model [Darvish and Crandall, 2001; Mendis et al., 1995; Miller, 1997; Nicolle et al., 2005; Prange and Margulies, 2002; Takhounts et al., 2003]. The large strain/high frequency behavior of brain tissue is newly added into this database [Pervin and Chen, 2009]. However, documented FE models (Table 2) have not been updated yet to reflect the new experimental data in predicting human head/brain behavior under blast conditions. Computational analyses of dynamic response of the brain under blast conditions only covered the low frequency response (
International Journal of Applied Mechanics Vol. 3, No. 4 (2011) 803–823 c Imperial College Press DOI: 10.1142/S175882511100124X
DYNAMIC RESPONSE OF BRAIN SUBJECTED TO BLAST LOADINGS: INFLUENCE OF FREQUENCY RANGES
MEHDI S. CHAFI∗ , SHAILESH GANPULE∗ , LINXIA GU∗,†,‡ and NAMAS CHANDRA∗ ∗Department
of Mechanical and Materials Engineering University of Nebraska-Lincoln Lincoln NE 68588-0656, USA
†Nebraska
Center for Materials and Nanoscience Lincoln NE 68588-0656, USA ‡[email protected]
Received 31 December 2010 Accepted 21 August 2011 Blast wave induced a frequency spectrum and large deformation of the brain tissue. In this study, new material parameters for the brain material are determined from the experimental data pertaining to these large strain amplitudes and wide frequencies ranging (from 0.01 Hz to 10 MHz) using genetic algorithms. Both hyperelastic and viscoelastic behavior of the brain are implemented into 2D finite element models and the dynamic responses of brain are evaluated. The head, composed of triple layers of the skull, including two cortical layers and a middle dipole sponge-like layer, the dura, cerebrospinal fluid (CSF), the pia mater and the brain, is utilized to assess the effects of material model. The results elucidated that frequency ranges of the material play an important role in the dynamic response of the brain under blast loading conditions. An appropriate material model of the brain is essential to predict the blast-induced brain injury. Keywords: Brain; hyper-viscoelastic material model; high frequency; finite strain; blast wave.
1. Introduction Complete understanding of mild traumatic brain injuries (TBI) induced by blast waves is challenging, due to the fact that currently no medical diagnostic tools could indicate the onset of the ailment [Hoge et al., 2008]. Finite element (FE) modeling has been widely used to predict the blast-induced brain responses and better understand the mechanism of TBIs [Moore et al., 2009; Moss et al., 2009; Taylor et al., 2009; Chafi et al., 2010; Ganpule et al., 2010; Grujicic et al., 2010]. The numerical predictions depends on the appropriate material characterization under blast loading conditions. ‡ Corresponding
author. 803
December 31, 2011 9:22 WSPC-255-IJAM S1758-8251 00124
804
M. S. Chafi et al.
Brain material is strain- and frequency-dependent [Bilston et al., 2001]. Blast scenarios require knowledge of brain tissue behavior over a large strain/high frequency range [Pervin and Chen, 2009]. Current experimental research has focused on the large strain/low frequency or small strain/high frequency behavior of the tissue, as summarized in Table 1. Some material models are based on hyperviscoelastic assumptions, i.e., a linear viscoelastic model in conjunction with a nonlinear hyperelastic model [Darvish and Crandall, 2001; Mendis et al., 1995; Miller, 1997; Nicolle et al., 2005; Prange and Margulies, 2002; Takhounts et al., 2003]. The large strain/high frequency behavior of brain tissue is newly added into this database [Pervin and Chen, 2009]. However, documented FE models (Table 2) have not been updated yet to reflect the new experimental data in predicting human head/brain behavior under blast conditions. Computational analyses of dynamic response of the brain under blast conditions only covered the low frequency response (
